Technical Field
The present disclosure relates to a fully balanced differential difference amplifier (FDDA or FBDDA) and to a device including the FBDDA amplifier.
Description of the Related Art
As is known, numerous types of circuits are used as front-end reading circuits for capacitive sensors. In particular, it is known to use FDDA or FBDDA amplifiers (Fully-balanced Differential Difference Amplifier), which are preferable when a high input impedance, a fully differential architecture, and unity or unit gain are required. A FBDDA amplifier of this type is shown by way of example in FIG. 1, and designated as a whole by the reference number 1. A capacitive sensor 2 belongs, for example, to a gyroscope, a pressure sensor, an accelerometer, a microphone, etc., and detects a variation of capacitance generated by a linear or rotational movement of the mobile parts of the sensor itself. The FBDDA amplifier 1 detects a variation of the inputs resulting from conversion of the desired physical quantity (e.g., pressure, rotation, acceleration, etc.) and generates at output a voltage proportional to said variation.
In the example of FIG. 1, the FBDDA amplifier 1 includes four input terminals 1a-1d and two output terminals 1e, 1f. The terminal 1a is an inverting terminal, and the terminal 1b is a non-inverting terminal. On the terminal 1b an input signal (voltage) Vin is present, including a voltage component generated by the capacitive sensor 2 and a fixed voltage component VCM. The voltage VCM, applied to the terminal 1b by a resistor 3, is a voltage for biasing the sensor 2, chosen according to the need for fixing the operating point of the sensor 2 (VCM is chosen, for example, in a range comprised between a supply voltage VDD and a voltage of a reference node, e.g., ground reference). The resistor 3 has a high value of resistance, for example 100 GΩ or higher. The voltage VCM is a fixed (d.c.) voltage and is further supplied to the input 1c of the FBDDA amplifier. In this way, the inputs of the FBDDA amplifier 1 are always biased at a common voltage, which is known. During use of the sensor 2, the varying (i.e., a.c.) input signal Vin is superimposed on the voltage VCM.
The signal on the terminal 1a is transferred onto the output terminal 1e. In other words, the output terminal 1e is feedback-connected to the input terminal 1a. Likewise, the signal on the terminal 1d is transferred onto the output terminal 1f so that the output terminal 1f is feedback-connected to the input terminal 1d. 
In a per se known manner, according to the operation of a FBDDA amplifier in voltage-follower configuration, on the output 1e the signal Vin/2 is present and on the output 1f the signal −(Vin/2) is present.
The embodiment of FIG. 1 guarantees good performance in terms of signal-to-noise ratio (SNR) but is deficient as regards the total harmonic distortion (THD) when the level of signal Vin increases. This is due to the fact that the operational amplifier is not connected according to a closed-loop configuration of a traditional type, and its inputs are not virtually connected together. Thus, in the presence of a high input signal Vin, the two differential pairs are markedly unbalanced, thus causing a deterioration of the linearity.
In this context, the linearity is referred to the differential-input pairs of the FBDDA amplifier, obtained by transistors. For small input signals (e.g., a.c. signals in the range between −150 mV and +150 mV, extremes included), the signal at output from the amplifier is substantially a replica, possibly amplified, of the input signal. Instead, for signals with a high peak value (e.g., a.c. signals having a peak value higher, in modulus, than approximately 200 mV), the transconductance gain of the two differential input pairs starts to differ markedly, thus generating a harmonic distortion of the output voltage signal (or differential output signal).
In order to overcome this disadvantage, it is known to use degeneration resistors 4 coupled to each differential-input pair (i.e., between 1a and 1b, and between 1c and 1d), as shown schematically in FIG. 2. This approach makes it possible to improve linearity of the output signal of the FBDDA amplifier, at the expense of noise that is the higher, the greater the degeneration resistances.
A further known solution, shown schematically in FIG. 3, envisages use of dynamic-biasing circuits 6 such that the current for biasing the differential pair is not fixed, but varies as a function of the input signal Vin. In this case, when the input signal Vin increases beyond a threshold value, a current is introduced into the differential pair. Even though this approach affords advantages in terms of input noise and harmonic distortion, it presents the disadvantage of requiring an excessive current consumption and a marked variability of the d.c. operating point.